^^

LIBRARY

OF THE

MASSACHUSETTS INSTITUTE

OF TECHNOLOGY

Structure of Individual Goals:

Implications for Organization Theory

Peer Soelberg

143-65

This paper was read at the 12th International Meeting of the Institute

of Management Sciences in Vienna, September 10, 1965. It may not be used

or reproduced without the permission of the author

.

Massachusetts Institute of Technology

September, 1965

RtCEiVED

JAN 4 1966 I

M. I. T. LIBRARJES ^

STRUCTURE OF INDIVIDUAL GOALS:

IMPLICATIONS FOR ORGANIZATION THEORY

Relevance of a Theory of Organizational Deci sions

I guess none of us would take issue with the statement that human choice

behavior is a central aspect of organizational life. Perhaps ten years ago

we might have found theorists who would have argued differently. But oppon-

ents of this view are surely hard to find today So accepting without further

commentary the importance of having available a sound theory of choice for

guiding descriptions of organizational behavior let us ask ourselves instead,

how adequate for that purpose are currently available decision theories?

An immediately noticeable characteristic of a disturbingly large number

of presumably organizational decision theories is that they are indistinguish-

(2)able from counterpart theories of individual choice behavior, even though

in its f lesh-and-bones reality we don't usually have much difficulty distinguish-

ing an individual from an organizational decision body.

Differentiation of Organizational from Personal Choice Theory

Yet it is not many years ago that it was heralded as a major accomplish-

ment by far-sighted theorists to have noted the rather obvious contrast in

social reality of a single person versus a group or an organization, and to

have begun to speculate about possible implications of this distinction for

(3)a more differentiated theory of organizational choice. But try as they

may even astute model-builders like H.A Simon seem to be having difficulty

weaning their organizational analyses from individualistic theoretical mother-

(4)hood . This paper examines some of the reasons why we should expect it

would be difficult to remove "the individual decision-maker" from theories

of organizational choice behavior.

The discussion below focuses on the notion of a decision "goal" -- the

means whereby Dm presumably discriminates among, is sometimes said to "prefer",

one choice alternative compared to a sub set of other, also availbale alternatives,

All theories of decision making, indeed any description of motivated behavior

in all areas of social science begins by imputing some sort of "goal",

"direction", or task orientation to whatever organisms it purports to describe.

-3-

Organizational versus Personal Goals

We begin by examining the question, how do organizational goals differ

from individual or personal goals? Consider first three types of answers

customarily given to this queries

1^. Traditional economic theory, Weberian bureaucratic theory,

and traditional management theories, suggest that there is no significant

difference betv;een an organization's goals and its individual participants'

"organizational role" goals. Even if differences might "hypothetically" be

shown to exist between goals of individuals and organizational objectives

the former have been excluded from relevance to organizational life, accord-

ing to the just-mentioned theories, by virtue of the existence of an "employ-

ment contract" between the organization and its members -- which presumably

secures member adherence to organization goals, sometimes called "contributions','

in return for some form of employment wage, a utility, or "inducements",

payment to said various members.

Whereas economists apparently trust individual employees to conform to

organization goals upon the mere paym.ent and receipt of an agreed-upon employ-

ment "price", bureaucratic and management theories add in a generous measure

of organization "control" as well, presumably thus assuring that employees

indeed will adhere to the organization's goals, except in exceptional, and

theoretically pathological, cases.

2. Traditional organization conLrol procedures have long appeared

unpalatably Orwellian to kindly humanistic theorists. Correspondingly, in

contrast to the old, hard "Theory X", recent more enlightened theoreticians

have suggested that personal adherence to organization goals may be achieved,

bynot necessarily, if at all authoritarian decree and control, but rather by

means of group discussion and consensus, with a fair measure of willingness to

(9)compromise supplied by the "official" organization. Nevertheless, the

outcome of such discussions among organization members is presumably still the

same, namely that a joint goal structure or an organizational preference ordering

will emerge, which then is to serve as the prescription for organizational

decisions to strive towards .

3. Along have come Cyert and March, among others, pointing out

that such ideal consensus among organization members is in fact impossible to

achieve, due in part to:

a the presence of bona fide, and simultaneously irreconcilable, conflicts among

different organization members' personal goals and preferences;

b. limits on human computational capacity for constructing and then utilizing

a complete preference order of (organizational) alternatives;

c_. communication barriers between even presumably cooperating parties; as

well as ,

d^. the common incidence of inconsistency and dynamics in members' own personal

goal structures, which makes a static description of even individual goals a

tenuous business at best. Furthermore, argue the authors quite persuasively,

the act of reaching member consensus, or definition of a complete joint organ-

izational preference ordering, is observably quite unnecessary for the perfectly

satisfactory operation of most every-day organizations.

Instead Cyert and March suggest that organizational goals should be viewed

as being a set of partial preference orderings, defined with respect to certain

frequently recurring, but semi-isolated, sets of alternatives which are ecoun-

tered separately by various organizational sub-units. These partially conflict-

ing organization goals are viewed by the authors as being institutionalized by

organizations in terms of

_i. standard operating procedures, or organizational "decision

programs";

ii a functional division of labor among organization sub-units, each

of which will presumably be pursuing somewhat different, at times

conflicting, organizational sub-goals; and

iii . budgetary resource allocations to the various organization sub-

units, which, due to an accompanying discretion for spending

such resources within often quite broad guide-lines, allow sub-

units to pursue partially conflicting goals simultaneously and

semi-independently

.

The Nature of Organization Goals in Received Doctrine

We want to keep in mind not only the postulated relationships of personal

to organizational goals in current choice theory, but also note the

exact nature of the types of goals assumed by these theories to be pursued

by hypothetical organizations:

1_. Economists, for example, seem willing to bend in almost any

direction to preserve their traditonal beliefs in the analytically tractable

"profit maximization" notion of organization goal. Quite recently, however,

Baumol did concede the possible existence of an organizational "sales" goal

as well -- which he (alas) immediately assumed got itself maximized by industrial

(12)firms, subject to a profit constraint. Baumol 's sales-goal modification

of classical economic profit theory is, in my opinion, unhappily restrained by

the author's refusal to carry his theoretical formulations beyond the notational

(13)niceties of traditional economist s function analysis.

Classical "management" theorists have been less mathematical in their

descriptions of organization goals, but are typically as insistent as their

economist counterparts that organizations do (or should) maximize dollar

(14)profit returns (over some, usually unspecified^ planning horizon).

Bureaucratic theorists unhappily find themselves somewhat at a loss compared

theyto economist collegues, as are often heard lamenting their lack of quantitative

numbers like profit and 'price to theorize with Instead hypothetical bureau-

cratic organizations are ofter said to maximize "public welfare", however that

^ (15)IS to be measured .

2. It is hard to infer, in general, just what "Theory Y" writers

consider to be the goals of organizations they describe. But this group of

theorists generally seems to focus heavily on organizational variables affecting

"employee welfare" and members' "personal growth," which they dutifully argue

ought to be increased substantially by present-day industrial organizations,

(16)presumably then subject to some sort of constraint on profit returns.

3. The Cyert and March theory claims industrial organizations strive

to achieve satisfactory performance on chiefly four different organizational

goals, in making pricing and production out-put decisions -- which traditionally

have been the only organizational quantities of much interest to analytical

economic theorists. Specifically these four organizational goals are: an

Inventory goal, a Production Stability goal, a Sales Quantity and/or Market-

share goal, and a Profit goal. With respect to each of these goals various

organization sub-units (departments) define for themselves certain explicit

"aspiration levels", which they then endeavor to achieve by executing sundry

decision routines ("programs"). The aspiration level goals are subsequently

revised "upward" or "downward" by each department in direction of the latter 's

recent "performance" with respect to each goal measure .

-8-

Implication of Received Doctrine

A common property of all above mentioned organization theories, stated

more or less explicitly by their respective authors, is that invoking one or

two or four exclusively organizational goals will suffice for explaining most

of the behaviors it behooves organization theorists to be interested in under-

standing or predicting.

Emerging Multi-dimensional Points of View

Yet there are voices heard advocating a more pluralistic view of (organ-

izational) decision goals, even from die-hard economist quarters. Georgescu-

Roegen for example stated the case for multi-dimensional, lexigraphically ordered

/IONutility as early as 1954. Later Chipman <Jhowed how vector utilities

could be handled axiomatically , by a symbolic notation at least as esoteric

as that of either Hicksian or Von-Neumann Morgenstern type analysis -- demon-

strating in effect classical Archimedean utility is merely the special case of

:or wl:

(20)

(19)lexicographic preference orderings for which the Axiom of Substitution

among goal-attributes is assumed to hold.

Shubik suggested four years ago that firms might constructively be viewed

as maximizing profits, but subject to a whole set of boundary constraints on

(21) (22)that goal. And Encarnacion quite recently scratched the surface of pos-

sible formalizations of what we might call "lexicographic satisf icing" behavior:

cases where Dm is observed to pay attention to a single one of a set of currently

non-satisfactory goals, or "problems", according to a (lexicorgraphic) priority

ordering among such goals -- such that, if and only if a higher order goal is

satisficed will Dm turn his attention to resolving the "next lower" goal on his

priority list .

-9-

Finally Charnes and Stedry just derived a set of interesting recommend-

ations for how a Dm should allocate his effort among multiple goals (actually multiple

problems, each problem defined by the same scalar goal measure), under an analytic-

ally tractable set of special assumptions about the technological structure of

(23)Dm s task environment.

Last year Simon finally drew the consequences of the "multi-dimensional"

(24)movement in organization-goal theory. And, since I would like to employ his

definitions as a point of departure for developing the theory further, let us

consider for a moment Simon's ideas in some degree of detail.

Simon's Definition of Organizational Goals

Simon notes a common property of most mathematical decision programs

that maximize, (or satisfice) a given objective function -- or no objective

function at all -- subject to a large set of constraints:

" ....a course of action, to be acceptable, must satisfy a whole

set of requirements, or constraints. Sometimes one of these

requirements is singled out and referred to as the goal of the

action. But the choice of one of the constraints, from many,

is to a large extent arbitrary. For many purposes it is more

meaningful to refer to the whole set of requirements as the (complex)

(25)goal of the action."

-10-

In order to define "organizational" goals Simon nov? simply draws the analogy

between the complex goal structure of a mathematical decision program and that

of an organizational role position:

"It appears convenient to use the term 'organizational goal ' to refer

to constraints, sets of constraints, imposed by the organizational

role ..."(2^>

Thus, if an organization is viewed as a collection of roles, each of which

is in turn subject to a large set of constraints on role incumbent's decision

behavior, the organization's "goals" become^ synonoraous with the totality of

all constraints acting on all organizational role positions.

Potential Criticism of Simon's Definition

Not a very elegant nor parsimonious definition of organization goals the

critic will immediately object. Nor, would I agree, will such a "total"

description of any organization's goal structure be particularly interesting to

possess, or even possible to obtain, in general. But "in general" we shall have

narrowed down dramatically the focus for our organizational analyses, being as

a rule interested in describing merely whatever constraint set happens to be

currently operable with respect to some specified sub-set of the organization's

members, or alternatively, with respect to some specifically defined organizational

"problem", set of problems, or other identifiable occasion for choice behavior.

And this, much more limited constraint set operating on sub-organizational role

behaviors will as a rule be considerably more tractable observationally , as well

as analytically.

-11-

Personal versus Organizational Goals Reconsidered

Having come this far let us now quickly summarize the answers given by the

four types of organization theories we have considered to the question we posed

initially, namely what is the nature of the relationship of personal to organ-

izational goals?

1_. Economists and classical management theorists view organizational

goals as being synonomous with the personal goals of the owner -entrepreneur

,

leaving little room in their theory for the personal goals of other employees,

save as such considerations might be "used" as effective means for increasing

the latters' (employer-defined) productivity.

2_. Neoclassical, more people-oriented management theorists view

organization goals (ideally) as being determined by some form of democratic

consensus among organization members -- as opposed to being handed down by

autocratic decree -- yet seem to theorize as if the democratic employees

thereafter agree to act in accordance with the agreed-upon organizational pref-

erence structure, thus excluding as it were conflict or exercise of personal

goals in organizational decision making.

_3 Cyert and March view organization goals as being a small number

of organization sub-unit aspiration level goals, existing entirely separately

from members' individual goals.

-12-

4. Simon views organization goals as a large number of constraints

on organizational role behavior, wherein a role-incumbent's personal goals are

relevant only as variables affecting the incumbent's decision to join, stay in,

or leave his role position in the organization: "Thus desire for power and

concern for personal advancement represent an intrusion of personal goals upon

(27)organizational role...", such that, according to Barnard and Simon's

inducements-contributions theory, once a member has joined an organization, his

role enacting behavior is entirely determined by official organization goals,

by thei e. role constraints, "and hence we do not need any motivational assumptions

beyond those of inducements-contributions theory to explain the ensuing role-

enacting behavior."

Conclusion

I disagree with all the above views of the role of personal goals in

organizational decisions. Let me now employ Simon's "search" theory of choice

in an attempt to explain why I disagree.

The Relevance of Role Incumbents' Personal Goals

Simon's "satisf icing-search" theory predicts that an organizational unit,

say an individual role-incumbent, will actively search for and evaluate altern-

ative solutions to his (well-defined) decision problem sequentially, in a one-

at-a-time manner. Each alternative will then be either ACCEPTED or REJECTED as

it appears on the scene, or as it is retrieved from memory. If and only if the

last alternative was REJECTED will Dm continue his Search for a more acceptable

alternative

.

-13-

Observations I've made of decision makers in action suggest that Dms tend

to make wide use of what we might call "alternatives generators", i e . of

environmentally powered methods whereby a large number of possible alternatives

get themselves paraded before Dm for his consideration, often with no more initial

"search" activity required on part of Dm than his willingness to continue to

(29)process, i.e. evaluate, generator-presented alternatives.

Secondly, alternatives are as a rule processed or evaluated by Dm in

parallel , i.e. several-at-a-time . And Dm does not usually terminate his

decision making as soon as an "acceptable", i.e. a merely non-constraint-

violating, alternative has been identified.

Thirdly, and this is the major, yet untested, hypothesis of this paper:

In most cases the "official" role constraints that are relevant bo whatever

organization problem a role-incumbent Dm may be working on are not sufficiently

constraining to specify for him uniquely which of a set of available , and

presumably all organizationally acceptable, decision alternatives should be

selected by our organization-Dm.

Stated a bit differently: A number of the potentially constraining role

prescriptions, which might or "should" have been operable on the decision de-

liberations of any one organization member, are negotiable, and will over a

period of time have been significantly modified by the patently "personal"

goals of the role incumbent in question.

-14-

In other words, I believe it is (will be) easily demonstrated that

role incumbents in most cases are asked to, allowed to, or able to exercise

considerable, and organizational choice-wise highly critical, personal discretion

in solving what are presumably "purely" organizational problems. I obviously

expect this to become increasingly the case as we wander up the hierarchy

of organizational authority. Even so I believe our theories have tended to

blind us to the opportunity for even quite low-level employees in an organization

to exercise considerable and quite often critical discretion in making what are

supposedly highly "programmed" decisions.

Contra-phrasing Simon I would then conclude: "Thus desire for power and

personal advancement (for example) represent personal goals which are of central

concern to organizational choice theory."

I would even go a step further and hypothesize as follows: If a role

incumbent is provided with anywhere near an opportunity to exercise organizational

discretion in such a way as to be compatible with or help achieve his own personal

goals or motives, then he will indeed do so. "Power corrupts, and absolute power

corrupts absolutely."

-15-

Implication for a Description of Organizational Goals

Drawing the consequences of this point of view we see that in order adequately

to describe the decision goals of an organizational sub-unit, or those of a single

decision-making role-incumbent, it will not be sufficient merely to indicate:

_1. the legal, organizational, inter-group, or inter-personal con-

straint sets operant on his role behavior for the decision

problem we are interested in studying.

We must also be able to specify:

2. what costs or personal sanctions are perceived to be attached

to the role-incumbent's violation of such constraints on his

behavior as we can identify; or conversely, what personal

rewards are thought to be forthcoming to him should he not

violate the organizational constraints in question.

And then we must adequately be able to describe:

3^ the personal goal structures or motivations of the particular

individual role-incumbents we are studying.

Measurement of such Personal-Organizational Goals

The first set of measurements I shall simply assume is presently empirically

feasible, obtainable say by means of quite direct interviewing procedures from

the organization members involved in whatever slice of the choice process we

are focusing on.

-16-

The second set of measures seems to be a good deal less public. Personal

costs, Dms ' private evaluation of potential sanctions, and possibility

estimates of "being caught" violating various organizational role constraints

are not likely to be explicitly formulated in advance of such considerations

actually influencing role incumbents' decision behavior. We might thus

expect not to be able to measure "perceived costs of constraint violation"

directly, say by means of graded questioning. On the other hand we might

be able to obtain a partial measure by identifying that class of constraints

which is unquestionably taken as fixed by certain role encumbents, as well

as the class of constraints taken to be impudently violatable, or negotiable,

under which identifiable organizational conditions.

Finally, with regard to the third and perhaps most intractable dimension

of the organization goal measure indicated above I have somewhat more elaborate

suggestions to make:

On the structure of Personal Goals

Perhaps in no field of social science has its practitioners managed

quite so successfully to cloak themselves in a mantle of mystique as in the

study of personal goals, or as the area is often referred to by psychologists:

"values", "motivation", or "personality". Sigmund Freud certainly hoped to

clarify our notions of psychic energy or personal motivation by elaborating

his "libido", "ego", and "super-ego" concepts. Yet Freud's apostles have

not succeeded in providing, at least this writer, much grounds for hope that

we will be able to interpret the psychoanalytic "personality growth" model

sufficiently operationally in an organizational decision context, to make

-17-

to beFreudian theories of much direct assistance to us in our study of personal

decision goals. But having once gone to some length in trying to apply

psychoanalytic theory in this respect, I have at least learned the following

lesson:

A theory of personal goals, to be useful for organizational analysis,

would hopefully demonstrate that:

A. Personal decision goals in most instances consisted of a very small set

of what we might call Primary goal-attributes, say about one to three

in number, which if satisfied by an alternative, beyond a certain high

level of attainment, would dominate all other goal considerations lexico-

graphically.

"Sex" is certainly a fine candidate for consideration in that

regard. In an occupational context "social recognition", or

simply "staying out of trouble" might well be considered.

"Fear of loss of job", "personal insecurity tovi;ard authority

figures", or even "creature comforts" seem a priori quite

reasonable, as well as, obviously, straight forward "desire

to do a good job", "thirst for political power", or 'bmbition

for a specific career position". Different factors might

well be Primary goal-attributes for different persons.

B. Such a small number of personal Primary goal attributes -- if such indeed

could be shown to exist -- were reliably inferable from organizational

behavior, say by means of careful, indirect interviewing techniques, and/or

by more direct observations of subjects on, or even off, their jobs.

-18-

A sufficient condition for obtaining reliable raeasuements of

individual goals in this respect would obviously be that n

(say four) clinicians arrived at the same personal "profiles"

for the same individual subjects, independently of each other.

But obtaining such a result in practice appears to be almost

too much to hope for at present.

C. The inferred personal Primary goal attributes were stable over each

role-incumbent's various organizational decision problems. Or else our

theory would also have to predict which Primary goals would be associated

with which or be evoked by what organization role contexts.

If such were the case -- that Primary goals indeed did turn

out to be stable over decision contexts -- then we should finally

have obtained for ourselves an empirically acceptable definition

of the much maligned "Self" or 'Personality" concepts currently

so popular among certain persuasions of organizational theore-

ticians .

D. Finally, Dm's personal Primary goals were (will need to be) translatable

by our theory into the particularized descriptor attributes relevant to

each specifically defined choice problem facing Dm in the exercise of

his organizational role

.

For example, "social prestige", when deciding if and where to

relocate his company's Research and Development Division, might

well be translated into Dm's being 'biased toward a Cambridge

location" -- whereas the same underlying "social prestige"

value gets translated into "bias toward time-sharing computer-

-19-

ization" when the same Dm is deciding if and how to revaimp

his Division's manner of cost accounting.

The Confirmation Theory of Decision Making

Having spent the better part of three years observing fairly sophis-

ticated decision makers solving problems in a variety of choice contexts I'd

say it is necessary, in order to describe choice behavior adequately, to dis-

tinguish sharply between at least three, perhaps four phases of the decision

(31)process, more specifically, between the:

_!. Problem Definition and Planning;

II Decision Design, Search, and Evaluation; and

III Choice Confirmation and Commitment Phases.

It would be much too lengthy to go into details below, so we shall have to

limit our discussion by focussing on some of the organizational implications of

this unfortunately somewhat elaborate theory of human choice. Consider each

of the above problem solving phases in turn.

I^. Problem Definition and Planning

Dm's initial recognition that indeed he, or the part of the

organization he identifies with, has a problem to solve -- and the subsequent

decision to allocate attention and resources to working ir^hat "area" of

the organization environment -- are extremely important, poorly understood,

and to date little studied processes of organizational decision making. There

-20-

is however little doubt in my mind that this phase of the choice process,

particularly in cases where such problem finding is highly discretionary,

provides occasions for individual role incumbents to exercise what are now unrec-

ognizedly wide powers of decision making, enabling them thus to influence

organizational resource allocation ir^ directions often highly compatible

(32)with their personal Primary goals.

Such opportunities for major influence on final choices, by individuals

who are not hecesarily high in the "formal" organizational power hierarchy,

exist to a large extent also in the subsequent phase of decision making, namely

at the Problem Definition and Solution Planning stage —particularly if the

problem found to be facing the organization is agreed to be "novel", or

thought to be particularly difficult to solve. Operationally, the first

part of this decision phase serves to transfrom an organizational sub-unit's

vague recognition that "a problem exists" into a Problem Definition in terms

of a. Similar Problem Types previously encountered by Dm and/or b^. a specifi-

cation of one or more "Ideal Solutions" Dm hopes to attain for such a Problem

Type. The latter part of this phase is then characterized by Dm's attempt to

derive a set of operational detail criteria, i .e . a Plan, which is then used

for generating and evaluating decision alternatives.

II. Decision Design, Search, and Evaluation

Above I already mentioned one implication for organization theory

viewingof our Dm's Search for Alternatives, not as a sequential, say aspiration-

level controlled exploration of the problem environment, but in terms of a

"parallel" evaluation process, where Dm in effects operates on a set of gen-

erated, and organizationally potentially acceptable, choice alternatives simul-

taneously (p .13) .

-21-

Similarly, during the Decision Design phase, my observations clearly

indicate that Dms in most situations evaluate each considered choice alter-

native by means of a series of Screening Stages, such that each new alter-

native is successively more thoroughly investigated, and simultaneously

evaluated, by Dm's stepwise collection of additional information about it --

which is then passed through an increasingly elaborate set of constraints or

goal attribute tests. These tests are as a rule defined along quite different

and usually non-compared goal dimensions.

Even more interesting to our discussion here is the following observation

regarding the actual choice process:

If an alternative violates one or more of Dm's "important"

constraints, what we might call Secondary Goals, then this is

usually a sufficient condition for his rejecting that alternative,

or at least for his attempting to redesign it. On the other hand,

a given alternative's non-violation of all Secondary and Primary

goals dimensions is by itself not generally a sufficient condi-

tion for Dm's settling for it as the solution to his problem.

However, should an alternative happen to s core "unusually good"

thison any Primary goal then together with the alternative's non-

violation of all other Primary goals and most Secondary goals,

does constitute a sufficient set of conditions for Dm's acceptance

(33)of it as his solution to the problem.

Furthermore, the data referred to above provides strong support for the

interpretation that each decision maker does indeed consult only one, two, at

-22-

most three Primary goal dimensions when making his final choice selections.

Whether these Primary goals attributes are also stable and can be predictably

translated from one organizational choice context to another, are obviously

rejectable empirical questions -- which I'm not prepared to answer at this

point -- but which, as indicated, constitute the obvious next hypothesis io be

examined by this type of research.

Ill Choice Confirmation and Commitment

However, having once identified for himself a Choice Candidate

alternative. Dm is by no means finished with his decision processing. He now

enters the next phase, which lent its name to this version of a decision theory;

namely his Confirmation process. Most Dms appearantly need to construct them-

selves "rational" arguments, decision rules in effect, -- a "utility weighting

function" if you please -- for why the particular alternative that has already

been implicitly chosen is indeed Dm's "best solution" to the defined problem.

Confirming a choice is at times a laborious, time-consuming, and often

quite an anxiety-producing process. I shall not enter into here just how

(34)Decision Confirmation takes place -- that has been done elsewhere, -- but

I would like to point out a few organizational implications of the existence

of such a decision making sub-process.

Most obviously the presence of need-for -Confirmation implies that the

taking of decision is a non-symmetrical and partially irreversible process.

Any given choice alternative will thus not be evaluated similarly before

versus after Dm has implicitly determined his choice for any given problem.

-23-

In general Dm's post-hoc Confirmation Decision Rule, i.e. his choice rational-

ization in effect, is usually not identical to, nor necessarily congruent with,

the usually much simpler manner in which Dm originally went about reaching his

implicit decision. Yet every new alternative that now domes along, after

Confirmation has begun, will now be subjected to Dm's "artificial" decision

rule, and will in fact in all but the rare case be summarily rejected, even

in instances where no visibly external costs have yet become attached to Dm's

(32)re-opening the problem for reconsideration.

In short, when decisions have reached the stage that they can be privately

"reasoned through" they are already well on the road to becoming ultra-stable

and well-protected by an explicdy, if only partially "true", decision rule.

The Confirmation hypotheses X7;ill be stronger in their effects, I would predict,

the more important the decision problem facing him is in Dm's personal life-

space. So concluding our discussion let me now point our at least one inter-

esting implication of the Choice Confirmation phenomenon for our theories of

organizational behavior.

A Final Implication of the Confirmation Decision Model for Organization Theory

Their choice of occupation or organization to work for is a very important

decision for most organizational participants. Confirmation theory would pre-

dict that "decision to participate" and the "decision to leave" are non-symmet-

rically such that most Dms ' will exhibit a tendency to want to stay with the

organizations they have chosen to work ir . Furthermore, as Festinger's Dissonance

theory woyld lead us to expect, Dm's explicable decision rule for why his par-

ticular choice of

-24-

organization is "best for him" -- provided only that his Primary goals are

not grossly violated meanwhile -- will become even more elaborate, and

Increasingly explictly communicable, after Dm has been working in his organ-

ization for a while. Indeed, I would hypothesize that Dm's explanation

for why he doesn't change organizations will adapt to meet even quite dramati-

cally changed organizational or environmental conditions, if Dm's Primary goals

are not grossly violated thereby.

Conversely, if Dm's Primary goals should become grossly violated, the

prediction is that Dm will then dream up all kinds of reasons -- e.g. will

add additional Auxilliary goal attributes to his goal structure -- for why

he wants to move away from his organization. Yet empirically such a direct

violation of one of Dm's Primary goals would seem to be a relatively rare

phenomenon for a Dm to experience in an organization which he nad orginally selected

patently for its outstanding value along those very goal dimensions.

Thus I conclude that organization theorists have been altogether too

much occupied with trying to explain "organizational mobility", and have not

yet paid enough attention to exhibiting the reasons for the amazing stability

of membership observable in most every-day organizations. Relatively speaking only

a small number of people actually get up and leave their organizations of their

own initiative I would predict. Of people that do leave a large number, indeed

the majority, I expect will be found to do so for organizationally "external"

reasons, i.e. due to changes iryfcheir personal lives that have little or

nothing to do with organization-influenced factors.

-25-

More specifically, I would guess that -- I haven't looked at relevant

data yet -- a majority of people in fact shift organizations either because:

a. their own prior, or post-hoc, career Plan, their family obligations,

or personal life cycle simply called for such an organizational

change to occur -- not that they were necessarily "unsatisfied"

with their ovm performance or the rewards of their organizational

role ;

b^. some un- searched -for job opportunity came along unexpectedly,

vjhich they "just couldn't afford to turn down" -- not that they

were necessarily "unsatisfied" with their present organizational

role and thus actively searching for new opportunities; or

c^. they were simply asked to leave the organization they were

with -- not that they were necessarily "unsatisfied" with

their present organizational role

.

Otherwise, as long as their Primary goals are not grossly violated, Dms

once working in an organization will tend to bend like rubber men -- a species

distinct from either economic or administrative man -- in adapting their

"organizational participation decision rule" to almost any subsequent violation

of their Secondary (and Auxilliary) personal goals.

In effect, contrary to inducements-contributions theory, we expect that

Dm will not be observed to shift organizations whenever "inducements" fall

below "contributions" -- particularly where "contributions utilities" are

defined to be the "inducement utility" of Dm's best presently perceived

(37)alternative, i.e. his current opportunity cost, and inducement 'utilities"

-26-

are measured by "weighting and summing" alternatives' value-attributes

according to the relative importance weight which Dm in his reported

decision rule claims that he is attaching to the sundry Value or goal

dimensions imbedded in his occupational choice.

Conclusion

Returning full circle then to our irtitial argument: lf_ the decision-

not-to-leave-the-organization is thus observed to be ultra-stable and well-

protected we are provided with yet another potent reason for believing tliat

most Dras indeed will seek expression and fulfillment of their Primary and

Secondary goals in the exercises of their organizational roles, whenever

opportunities are found to do so v;ithout incurring prohibitive personal

costs by directly violating well-defined organization expectations or con-

straints. The plainly personal goal structure of key individual role incum-

bents for this reason becomes an important ingredient to study even for

theorists presumably interested in understanding and predicting "purely"

organizational choice behavior. This is surely not a starting conclusion

for us to arrive at, but it points out a direction which too long has been

deserving attention and conceptualization by analytical organization theorists

NOTES

1. Yet see E Schein, who in his Organizational PsycholoRy , Englewood

Cliffs, N J , 1965, manages to get along without a single reference

to human choice or organizational decision processes.

2. The traditional economic theory of the firm is of course an outstanding

example, for examples see Note 6.

3. For example, R.M Cyert and J G March, "A Behavioral Theory of Organizational

Objectives," in Modern Organization Theory , M- Haire (ed . ) , New York, 1959.

4. From "...the basic features of organization structure and function derives

from the characteristics of human problem solving processes and rational

human choice" (Organizations , 1958, p. 169), to "It appears convenient to

use the term 'organizational goal' to refer to constraints ... imposed by

the organization, that have only an indirect relation with the personal

motives of the individual who fills the role." "On the concept of

Administrative Quarterly of Organizational Goal," £, 1964, p.l.

_5. "Decision maker" is hereby abbreviate to "Dm".

6^. See for example, G. J. Stigler, The Theory of Price , New York, 1946,

pp. 20-38; J. M. Henderson and R. E. Quandt , Microeconomic Theory , New

York, 1958, pp. 42-75; or R. Dorfman, The Price System , Englewood Cliffs,

N. J. , 1964, pp. 14-72 .

_7. See for example, M. Weber, The Theory of Social and Economic Organization ,

translated by T- Parsons, Oxford, 1947, pp. 324-329; D Waldo, The

Administrative State , N. Y., 1948, pp. 76-88; or F. M. Marx, The Adminis -

trative State Chicago, 1957, pp. 37-45.

-2-

8^. See for example L. H. Gulick and L. Urwick (eds.) Papers on the Science

of Administration , New York, 1937; pp. 49-78; or P.F. Drucker , The Practice

of Management , New York, 1954, pp. 351-369.

9.. D. M. McGregor, The Human Side of Enterprise . New York, 1960, pp. 124-131.

10 R. M Cyert and J. G. March, The Behavioral Theory of the Firm , Englewood

Cliffs, N.J., 1963, Chapter 3.

11 . See for example K.E. Boulding and W. A. Spivey, Linear Programming and the

Theory of the Firm , New York, 1960, pp. 1-17.

12 W. J. Baumol, Business Behavior, Value, and Growth, New York, 1959, pp. 45-53

13 W. J. Baumol, "On the Theory of Expansion of the Firm", American Economic

Review , 52 1962, pp. 1078-87.

14 • p. F Drucker ££. cit .

15 M. Weber, £2. cit . See also T. Parsons, Social Structure and Personality ,

London, 1964, pp. 193-208.

16 See for example C. Argyris, Integrating the Individual and the Organization,

New York, 1964, pp. 146-163.

17 . Cyert and March, 0£. cit , Chapter 8.

18 . N. Georgescu-Roegen "Choice, Expectations, and Measurability" Quarterly

Journal of Economics , 68, 1954, pp. 503-34.

-3-

19 A lexiographic preference order derives from an underlying multi-

dimensional goal structure with zero rate of substitution and linear

hierarchical dominance between dimensions. In effect an alternative is

preferred to another if it "is better" than the latter on the hierarchically

top-most goal attribute, regardless of other alternatives' relative

scores on any other goal-attributes. Only in case of indifference withgoal

respect to the top-most will the next, hierarchically lower goal -dimension

be consulted to break the tie.

Opponents of the analytic complications introduced by assuming

lexicographic preference orderings object that only in extremely rare cases

v>7ill third, fourth dimensions, etc., be needed to break ties among altern-

atives As a compromise Banerjee suggests our being openminded enough at

least to accept two-dimensional lexicographic scales as a reasonable

analytic tool for economists to make use of, particularly for cases involving

stochastic uncertainty of choice consequences (D. Banerjee "Choice and

Order: or First Things First ," Econoraica, 31, 1964, pp. 158-167.)

20. J.S. Chipman, "The Foundations of Utility" Econometrica . 28, 1960, pp. 193-224,

21 M. Shubik, "Objective Functions and Model of Corporate Maximization",

Quarterly Journal of Economics ", 75, 1961, pp. 347-375.

22 . J. Encarnacion, Jr., "Constraints and the Firm's Utility Function",

Review of Economic Studies , 31, 1964, pp. 113-120.

23 . A. Charnes and AC. Stedry, "Search-Theoretic Models of Organization

Control of Budgeted Multiple Goals" Management Sciences Research Report

tig . 32

,

Carnegie Institute of Technology, 1965.

24. H. A. Simon "On the Concept of Organizational Goal", Administrative

Science Quarterly , 9, 1964, pp. 1-22.

25. Ibid. p. 7.

26.. Ibid. p. 21.

26a. March and Cyert appcarantly believe that sub-units' goals have arisen at

some point in organizational history from "coalitions" forming among

organization members, although the authors leave description of the process

to the reader's imagination. op . cit . , p. 29-32.

27 . H. A. Simon, 0£. cit . p. 12.

28. Ibid. p. 11 .

29 . p. Soelberg, First Generalized Decision Process Model: GDP-I , forthcoming.

30 . For example, even with a highly mechanized decision rule such as the one

installed in a paint factory by C. Holt, F Modigliani, J. Muth , and H A.

Simon (Planning Production, Inventories, and Work Force , Englewood Cliffs,

N J , 1960 did not prevent production schedulers from exercising quite a

(dysfunctional) bit of discretion in implementing the rule.

31 . P. Soelberg, ££. cit .

r

32 . H. A. Simon, in personal conversation, August 30, 1965.

33 . P. Soelberg, o£. cit .

34. Ibid.

-5-

35 . A process Leon Festinger , among others, believes is characterized by a

certain amount of "dissonance reduction," Conflict, Decision, and Dissonance ,

Stanford, 1964.

36. Some of V. Vroom's data, referred to in a recent working paper, 'Organizational

Choice, a Study of Pre- and Post-Decision Processes" Carnegie Institute of

Technology, 1965, is also compatible with this interpretation.

37 . J. G. March and H. A. Simon, 0£. cit . , pp. 84-88.

m^^"^

FF^ |3'7i

Date Due

RG S'89

MIT LIBRARIES

3 TOflO DD3

1L( l-dp^

6TT Sfl?

UIT LIBRARIES

/^-^ '6$

3 TDflD DD3 fiTT ST5-W'TJr'A'*.*.'' l^s.. .„ . ,. wAVA -^ WV

3 TDflD DD3 fiT'i t.D3HD28 M.I.T. Alfred P.Sloan

syjA^ .mUiU School of Management' Nos.lU2-65Working Papers.

3 TOaO DD3 ATT bll N0S.I5I-65

MIT LIBRARIES

MIT LIBRARIES

3 ^DflD QD3 flbfi b24

iM-/-fcf

lllllllllillllllniill'i'll'

3 TD60 003 fi=lT bEI/

3 lOfi 003 TOO

H%.&<

153B^SEI*E^^

SIGN this card and present it with book

3 TOfiO 003 fibi 135 at the CIRCULATION DESK,MIT LIBRARIES

1':'̂d'fe^'

60 003 TOO 136 MASSACPIUSETTS INSTITUTE OF

III ,,<- TECHNOLOGY LIBRARIES

II

\^t^^^

DEWEY LIBRARY

TO

3 TOaO 003 6

y